This is a keygen to solve the weasel keygenme challenge. It is based on the GPU version of libFES available here licensed under MIT.

To compile, open a new Python3 Colab Notebook and follow the steps below. Ensure that the project Hardware Accelerator is set to GPU. Alternatively, use this notebook and "Open in Playground" to run.

! git clone https://github.com/extremecoders-re/weasel-keygen
! chmod +x ./weasel-keygen/keygen.py
! chmod +x ./weasel-keygen/fix.py
! chmod +x ./weasel-keygen/gen_kernel.py
! cd weasel-keygen && make guess

To generate a password, simply run keygen.py. To make the process faster, a part (10 bits) of the password is randomly generated while the other half (40 bits) is calculated. As a result, occasionally it may fail to generate a password, in that case retry. Examples

! cd weasel-keygen && ./keygen.py 0xec
Username:  0xec
Password: YBG3T-6EWUG

! cd weasel-keygen && ./keygen.py kao
Username:  kao
Password: QL9UB-LTKYY

! cd weasel-keygen && ./keygen.py extremecoders
Username:  extremecoders
Password: RAGWC-XW6VP

You can also run the keygen on Kaggle. https://www.kaggle.com/extremecoders/weasel-keygen?scriptVersionId=27419227

• An adaption of the crossbred algorithm for solving multivariate quadratic systems over F2 on GPUs
• Implementing Joux-Vitse’s Crossbred Algorithm for Solving MQSystems over F2 on GPUs